Find the best datasets for your research question or project.

Discover Data @ Duke is a collaborative effort to provide a one-stop shop for Duke investigators to learn about and access a wealth of data resources.

CTSI and the DHTS Analytics Center of Excellence Research Solutions Team have partnered to connect investigators to local, regional, and national datasets that are crucial for scientific and clinical research.

Interested in learning more or starting a project using clinical data? Get in touch with our data experts.

DiscoverData @ Duke Offers:

Consultation Services

Data Query Tools

  • DEDUCE—query de-identified patient data from the Duke Health system. To submit a DEDUCE request, visit the Customer Portal.
  • i2b2uery EHR data from the Duke EHR and/or the EHRs of members of national network of CTSA institutions (aggregate counts only) via the ACT Network

Clinical Data Research Networks

Get patient population counts to determine feasibility and find collaborators for local and multi-site studies and clinical trials.

  • STAR CRN/PCORnet—regional and national clinical data network with potential access to a patient population of 66 million
  • ACT Network—national network of more than 39 CTSA institutions (18 more currently staging)
  • Carolinas Collaborative—regional network of premier institutions in North and South Carolina
  • Learn more about CDRNs and request a consult with our CDRN team

Datasets and Resources

  • Discover Data @ Duke Catalog—a catalog of data resources that can help you answer your project or research question
  • National Datasets—varied population health datasets to support research
  • Duke SOAR Data Inventory—a collaboration to open clinical research data for the benefit of the broader research community
  • Population Health DataShare—giving Duke researchers access to electronic health from different sources to generate new insights into health and healthcare. Provided by the Duke Department of Population Health Sciences
  • MURDOCK Study—MURDOCK Data Assets include 430,000 biological samples [serum, plasma, whole blood (DNA), paxgene (RNA) and urine] from 12,526 study participants in the MURDOCK Study Community Registry and Biorepository. The Community Registry includes self-reported data from baseline and annual follow-up questionnaires. Baseline characterization of the cohort included participant identification, contact information, primary physician; date and place of birth; demographics; self-reported health history encompassing 34 disease domains and related procedures; medications and supplements, dietary and physical activity; sleep, tobacco and alcohol use and selected PROMIS participant-reported outcomes domains, socioeconomic data, vital signs (height, weight, and waist circumference). Each participant is geospatially mapped at their home street address level to enable attribution of publically available social, economic, and natural and built environmental features to the individual using geospatial information systems (ACS and US Census). Participants were also consented for ongoing access to their electronic health records.